A scalable and backwards compatible transition solution for ID/Locator separation architecture

The scalability of the Internet routing system suffers from the overloaded semantics of lP addresses. The IP address of a node means not only the identifier but also the locator. Many researchers propose the idea of separating the Internet into two independent evolving spaces, the identifier space and the locator space. However, this architecture is quite different from the Internet we are using, so a good transition plan is crucial for the further deployment of the ID/Locator separation architecture. In this paper, we propose a new compatible method, in which the nodes in the traditional network and the nodes in the access network of the ID/Locator separation architecture can communicate with each other. And we give a detailed analysis of the method. We show that, besides that the method supports the communication between the end node in the traditional network and the end node in the ID/Locator separation architecture, the method does not take extra delay and has good scalability. Using this method, the architecture can be incrementally deployed, and offers security, multi-homing, and mobility benefits even to early adopters.